Recent in depth molecular analyses of human malignancies, through projects such as The Cancer Genome Atlas (TCGA), have revealed numerous frequently identified mutations; however, only a subset of these actually contribute to the development of particular cancers. The malignant phenotype is often the result of synthetic genetic interactions between multiple genomic and epigenomic aberrations. As such, subsets of tumors have specific co-occurring mutations or genomic alterations that cooperate in a co-dependent manner. The goal of this application is to identify the critical co-dependent molecular pathways that cooperate with known driver genomic alterations in oral squamous cell carcinoma (OSCC), one of the most frequent malignancies of the head and neck. The insight gained will, in turn, provide a platform for novel drug discovery and/or rationale for the investigation of novel combinations of existing drugs. Aim 1 will use sophisticated new bioinformatics algorithms developed by the Gevaert lab to integrate mutation and copy number alteration data with DNA methylation and gene expression data in OSCC TCGA data sets. These algorithms will be used to predict, with high probability, candidate genetic interactions among heterogeneous OSCC tumors and to identify master regulators of gene modules that are related to particular biologic processes, such as metastasis. In Aim 2, candidate gene interactions and master regulators will be validated by the Sunwoo lab using next generation in vivo synthetic lethality assays, using patient-derived xenografts to more closely reflect the primary tumor. Candidate master regulators of metastasis will also be evaluated using in vivo assays. In Aim 3, the Kuo lab has adapted their experience in culture and oncogenic transformation of gastrointestinal 3D air-liquid interface primary organoid cultures to OSCC. Accordingly, our validated wild-type oral mucosal organoid protocols will be used to introduce co-occurring mutations and gene alterations into wild-type human and mouse oral mucosa tissue to functionally validate the oncogenic activity and multigenic transforming synergy of putative OSCC genes from Aims 1 and 2. In Aim 3, the 3D organoid culture approach will also be used to grow primary human OSCC tumor organoids directly from surgical samples, for in vitro chemosensitivity testing, correlation against exome sequencing mutational status and shRNA/sgRNA-based gene validation. This bi-directional strategy of (1) targeting co-occurring mutations in patient-derived xenografts and primary tumors and (2) introducing co-occurring mutations into normal oral mucosa will provide important insight into our understanding of the synthetic genetic interactions in OSCC. Further, the functional and genetic data from Aims 2 and 3 will be channeled back to Aim 1 to continuously update the bioinformatics models.